validation pattern · L2ET Research Method Signal
Dimensional analysis is a low-cost falsification filter for theoretical models
Before numerical or symbolic complexity, proposed relations should satisfy unit balance, limiting cases and dimensionless-regime checks.
confidence: high
action: apply
score: 85
Theoretical Physics, Frontier and Theoretical Physics, Scientific AI
tools/dimensional-analysis-checkerframeworks/theoretical-physics-research-protocol
open related instrument →
validation pattern · L2ET Research Method Signal
Sensor fusion failures often begin with time sync, calibration and covariance drift
Fusion outputs should be checked against clock drift, calibration age, sensor disagreement, bias and covariance discipline.
confidence: high
action: apply
score: 85
Robotics, Scientific AI
tools/sensor-fusion-consistency-checkerframeworks/robotics-autonomy-assurance-model
open related instrument →
reproducibility pattern · L2ET Research Method Signal
Scientific AI credibility depends on reproducibility records
Data lineage, code availability, environment locking, seeds, uncertainty reporting and external validation determine how far a result can travel.
confidence: high
action: apply
score: 92
Scientific AI, Medical AI, Frontier and Theoretical Physics
tools/scientific-reproducibility-scorecardframeworks/scientific-reproducibility-model
open related instrument →
validation pattern · L2ET Research Method Signal
Medical and imaging AI requires evidence discipline before operational claims
Research-oriented medical AI discussion should separate dataset provenance, subgroup behavior, uncertainty, external validation and workflow fit.
confidence: high
action: apply
score: 90
Medical AI, Scientific AI
tools/scientific-ai-validation-checklistframeworks/scientific-reproducibility-model
open related instrument →
formalization pattern · L2ET Research Method Signal
Formal methods need stable interfaces, invariants and failure semantics
A system is more ready for formal specification when state variables, forbidden states, invariants, interfaces and environment assumptions are explicit.
confidence: high
action: review
score: 83
Robotics, Aerospace Engineering, Scientific AI
tools/formal-specification-readiness-checkerframeworks/robotics-autonomy-assurance-model
open related instrument →